ibrd & ibrdf not terminating on EOI line, timeout EABO

[3TMQ]

GPIB-USB-HS    Interactive Control

Reads with ibrd or ibrdf are terminating with a time out EABO, and ignore the EOI line.

if ibrd is set for the exact size (eg 21), it terminates correctly.

using

  ibdev 0 03 0  12  1 0

How dor I set to termiante on EOI line alone, so I don't have to know the size of the message (or termination character)?

[Dennis_Knutson]

What's the instrument? Is it actually 488.2 compliant? Check the manual to see if there is a configuration setting.

[3TMQ]

Hi Dennis,

This particular instrument is an old HP5335A counter, but before this I am sure I had the same problem capturing plots with a HP8713.

The only solution was to set a timeout, and ignore the EABO.

I guess I could set an EOS character, but I should not have to do that, and I am trying to understand why I don't see the EOI on an ibrdf or ibrd with a big buffer.

e.g. Here is what I get when the buffer is too long (or ibrdf):-

ud0: ibrd 22
[c100] ( err timo cmpl )
error: EABO
count: 21
53 20 20 2b 33 2e 35 35     S + 3 . 5 5
32 32 34 32 35 36 31 45     2 4 2 5 6 1 E
2b 30 30 0d 0a              + 0 0 . .

But if I happend to know that exactly 21 char ar sent, I get:-

ud0: ibrd 21
[0100] ( cmpl )
count: 21
53 20 20 2b 31 2e 34 30     S + 1 . 4 0
33 38 32 35 39 39 33 45     3 8 2 5 9 9 3 E
2b 30 30 0d 0a              + 0 0 . .

And with ibrdf <filename>, you do not get to specify the size, so the only way of ending appears to be TIMO.

I am sure I am missing something simple in the setup to enable the EOI to work with ibrdf, but I justy can't find it.

the steup is:-

gpib0: ibdev 0 16 0 12 1 0

ud0: ibwrt "wa1 sr1 re"
[0100] ( cmpl )
count: 10

My goal this time was to write a script to to to wait for the measurement then capture readings from a pair of these counters.

Many thanks

Michael

[Dennis_Knutson]

The eoi is sent by the instrument and if the instrument does not send it, the instrument is defective, does not support it, or is not configured for it. If all that it supports is something like cr/lf, you need to configure your read to terminate on that.

[nyc_(is_out_of_here)]

---

@3TMQ wrote:

This particular instrument is an old HP5335A counter, but before this I am sure I had the same problem capturing plots with a HP8713.

---

Check to see if this instrument requires gpib re-addressing.

I seem to recall using one of these at a job ten years ago.

[3TMQ]

Hello NYC and Dennis,

Many thanks for your replies.  I'm still trying to get to the bottom of this.......

1/

I'm not sure how to check if this instrument "needs" readdressing - it didn't come up in a search of the HP5335A manual, but that could be a .PDF or scanning thing.

I entered the following into Instrument controll to try it out:-

gpib0: ibdev 0 16 0 12 1 0

ud0: ibconfig ibcreaddr 1
[0100] ( cmpl )
previous value: 0

ud0: ibwrt "fn2 wa1 sr1 re"
[0100] ( cmpl )
count: 14

Made a measurement 

ud0: ibrdf x
[c900] ( err timo rqs cmpl )
error: EABO
count: 21

So still get the error on reading the result to a file
Made another measurement

ud0: ibrd 22
[c900] ( err timo rqs cmpl )
error: EABO
count: 21
53 20 20 2b 31 2e 30 37    S + 1 . 0 7
30 39 38 33 34 39 39 45    0 9 8 3 4 9 9 E
2b 30 30 0d 0a             + 0 0 . .

And also errors passing buffer greater than the exact length

So I don't know if readdressing is a problem or not.  It is certainly not the whole solution.

I do not understand what "readdressing" is either.......

2/

The HP5335A uses an MC68488 device to interface to the HPIB bus.

The LEOI line is connected via a tristate bidirectional buffer to the EOI line (pin 5), so I expect the instrument supports EOI in the bus transactions.

I assumed that because an ibrd 21 [i.e. with the exact buffer size] completes OK, then that indicated the GPIB-USB-HS was seeing the EOI line toggle.

Perhaps I am wrong on this last point.

Many thanks

Michael

[3TMQ]

Helo NYC

OK, thanks for previously posting your good explination of re-addressing

http://forums.ni.com/t5/Instrument-Control-GPIB-Serial/repeat-addressing-gpib-readdressing/m-p/32423...

I just found it 🙂

I'm not specifically setting 488.2 in "Instrument Control" so I expect it was not a factor anyway.

Cheers

Micahel

[nyc_(is_out_of_here)]

I was working with the HP5334A, and I had no problems with reading from the instrument.

Never used ibrdf( ).

I don't recall if it was this counter or something else which required the GPIB re-addressing.